Homebrew PCBs

I've just made another PCB with a footprint for a Hirose 50 way DF12
connector with 0.5 mm lead spacing, it turned out very well. The DF12
pads are 0.38x1.00 mm.

The PCB is double-sided with a continuous ground plane on one side.

I used my usual photo-etch technique with artwork printed onto JetStar
film using an Epson 870 printer.

I've uploaded a pic of part of the PCB to Photos, in my 'Leon's' folder.

Leon

Hi Leon,

>I've just made another PCB with a footprint for a Hirose 50 way DF12
>connector with 0.5 mm lead spacing, it turned out very well. The DF12
>pads are 0.38x1.00 mm.
>  
>
That's very impressive! It's pretty much what I'm aiming for, but I just 
can't seem to get my process right. It's good to see some results in 
context: I often find that people advocating one method or another 
without revealing the level of complexity (principally trace and space 
width) that they're attempting.

I'd like to be able to make double-sided boards with 8/8 traces and 
spaces but this seems problematic. I'd be interested to hear from anyone 
who's done this! My two biggest problems are contamination of the board 
while handling it between stages of the process, and alignment between 
the two layers. I need aligment errors of less than 1mm, and I'm not 
sure that this is achievable in a DIY set-up. At the moment I drill 
guide holes before exposure and use PCB pins to tack each side of 
artwork in place, in turn, before taping the artwork to the board and 
placing it in my UV box.


Stuart

----- Original Message ----- 

From: "Stuart Wallace" <stuartw@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Tuesday, June 13, 2006 1:00 PM
Subject: Re: [Homebrew_PCBs] My best fine-pitch PCB so far


> Hi Leon,
>
>>I've just made another PCB with a footprint for a Hirose 50 way DF12
>>connector with 0.5 mm lead spacing, it turned out very well. The DF12
>>pads are 0.38x1.00 mm.
>>
>>
> That's very impressive! It's pretty much what I'm aiming for, but I just
> can't seem to get my process right. It's good to see some results in
> context: I often find that people advocating one method or another
> without revealing the level of complexity (principally trace and space
> width) that they're attempting.

What are you printing the transparencies on?

>
> I'd like to be able to make double-sided boards with 8/8 traces and
> spaces but this seems problematic. I'd be interested to hear from anyone
> who's done this! My two biggest problems are contamination of the board
> while handling it between stages of the process, and alignment between
> the two layers. I need aligment errors of less than 1mm, and I'm not
> sure that this is achievable in a DIY set-up. At the moment I drill
> guide holes before exposure and use PCB pins to tack each side of
> artwork in place, in turn, before taping the artwork to the board and
> placing it in my UV box.

8/8 is feasible, but I prefer using 10/10 or 12/12. That board of mine has 
12 mil tracks. I have done one or two double-sided boards with tracks on 
both sides, by making a sort of envelope of the two transparencies with 
masking tape. The registration was slightly off (better than 1 mm, though) 
but I was able to assemble the boards OK.

Leon

8/8 is easy (& good aligment, say 0,1mm) if you use graphic film as artwork.
  
Stuart Wallace <stuartw@...> wrote:
          Hi Leon,

>I've just made another PCB with a footprint for a Hirose 50 way DF12
>connector with 0.5 mm lead spacing, it turned out very well. The DF12
>pads are 0.38x1.00 mm.
> 
>
That's very impressive! It's pretty much what I'm aiming for, but I just 
can't seem to get my process right. It's good to see some results in 
context: I often find that people advocating one method or another 
without revealing the level of complexity (principally trace and space 
width) that they're attempting.

I'd like to be able to make double-sided boards with 8/8 traces and 
spaces but this seems problematic. I'd be interested to hear from anyone 
who's done this! My two biggest problems are contamination of the board 
while handling it between stages of the process, and alignment between 
the two layers. I need aligment errors of less than 1mm, and I'm not 
sure that this is achievable in a DIY set-up. At the moment I drill 
guide holes before exposure and use PCB pins to tack each side of 
artwork in place, in turn, before taping the artwork to the board and 
placing it in my UV box.

Stuart



         

 __________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

[Non-text portions of this message have been removed]

On Tue, 13 Jun 2006 12:50:48 +0200, leon_heller  
<leon.heller@bulldoghome.com> wrote:

> I've just made another PCB with a footprint for a Hirose 50 way DF12
> connector with 0.5 mm lead spacing, it turned out very well. The DF12
> pads are 0.38x1.00 mm.
> The PCB is double-sided with a continuous ground plane on one side.
> I used my usual photo-etch technique with artwork printed onto JetStar
> film using an Epson 870 printer.
> I've uploaded a pic of part of the PCB to Photos, in my 'Leon's' folder.
> Leon


I've done 0.5 with toner transfer and it worked well also. The pitch is  
not yet the problem, but getting the signals away from pads in the middle  
like BGA you can't do that without THP. Also it can be _very_ hard to  
observe the suggested track length with the new high speed ICs where a  
single via is like a huge inductor.

ST

>> Leon wrote:
>> I've just made another PCB with a footprint
>> for a Hirose 50 way DF12 connector with 0.5
>> mm lead spacing, it turned out very well.
>> The DF12 pads are 0.38x1.00 mm.

> Stuart wrote:
> <snip>
> I'd like to be able to make double-sided
> boards with 8/8 traces and spaces but this
> seems problematic.

8/8 thou should be doable with photo as long
as you use pre-coated boards and have a good
light source and photo tool.

Most people say vellum works better than
transperancies.   Laser-star and jet-star
some people talk of are special vellum that
is meant to be "specialy formulated for laser
and ink jet printers respectivly.

I just use common garden variety cheap
vellum.

If you want to do the occasional 8/8 part of
a board then any laser printer or inkjet will
do.  If you want to reliably get a WHOLE PCB
that is choccas full of 8/8 stuff then a
photo tool with good density is a must.

If you want to get down to 6/6 or 4/4 then
you HAVE to go to a local print shop and ask
them to do a photo tool for you on their
typesetting machine.

> I'd be interested to hear
> from anyone who's done this! My two biggest
> problems are contamination of the board while
> handling it between stages of the process,
> and alignment between the two layers. I need
> aligment errors of less than 1mm, and I'm not 
> sure that this is achievable in a DIY set-up.
> At the moment I drill guide holes before
> exposure and use PCB pins to tack each side of 
> artwork in place, in turn, before taping the
> artwork to the board and placing it in my UV
> box.

Everyone has thier own trick for lining up DS
boards.

Mine is to get an L shaped peice of FR4 with
double sided tape on it.  I then use a clear
sheet of perspex of similar thickness in
place of the final pre-coated PCB.  Line this
up ontop of a light box by eye.  Then slip in
the real PCB to do the exposure.

You should be able to get things lined up
almost an order of magnitude better than 1mm
just by eye on your first attempt :D

Leon Heller wrote:

>What are you printing the transparencies on?
>  
>
I'm using a Canon i560 to print on IBM inkjet transparencies. I've tried 
a few different types of easily-available transparencies (the sort of 
stuff that places like PC World stock) and haven't noticed much of a 
difference between the types. I've made a number of test boards 
containing designs with 8/8 rules -- in each case the results haven't 
been nearly good enough to be usable. Something is definitely badly 
wrong with my process! I struggle to fully develop the boards without 
mechanical assistance (generally rubbing at the boards with cotton buds) 
-- this invariably leaves contaminants on the board, which in turn 
causes etching problems. Etching is in any case very uneven. This seems 
to point to either bad artwork or insufficient exposure.

I expose the boards (6" x 4" DS FR4 stock) about two inches above two 
15W UV tubes for three minutes. I'm extremely nervous about 
over-exposure: perhaps too nervous!

I'd love to be able to make 8/8 double-sided boards at home; the reduced 
turnaround time would really help some projects. I suspect, however, 
that DIY PTH may be beyond my abilities. I'm getting ready to send 
artwork over to Olimex in order to get some proto boards made up, but I 
have my doubts: my requirements seem to be at the absolute limits of 
what they offer -- 8/8, lots of fine-pitch high pin count SMDs, and so 
forth -- so I'm not sure whether the results will be any good. Still, it 
doesn't cost much to find out!


Stuart

----- Original Message ----- 

From: "Stuart Wallace" <stuartw@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Monday, June 19, 2006 12:09 PM
Subject: Re: [Homebrew_PCBs] My best fine-pitch PCB so far


> Leon Heller wrote:
>
>>What are you printing the transparencies on?
>>
>>
> I'm using a Canon i560 to print on IBM inkjet transparencies. I've tried
> a few different types of easily-available transparencies (the sort of
> stuff that places like PC World stock) and haven't noticed much of a
> difference between the types. I've made a number of test boards
> containing designs with 8/8 rules -- in each case the results haven't
> been nearly good enough to be usable. Something is definitely badly
> wrong with my process! I struggle to fully develop the boards without
> mechanical assistance (generally rubbing at the boards with cotton buds)
> -- this invariably leaves contaminants on the board, which in turn
> causes etching problems. Etching is in any case very uneven. This seems
> to point to either bad artwork or insufficient exposure.
>
> I expose the boards (6" x 4" DS FR4 stock) about two inches above two
> 15W UV tubes for three minutes. I'm extremely nervous about
> over-exposure: perhaps too nervous!
>
> I'd love to be able to make 8/8 double-sided boards at home; the reduced
> turnaround time would really help some projects. I suspect, however,
> that DIY PTH may be beyond my abilities. I'm getting ready to send
> artwork over to Olimex in order to get some proto boards made up, but I
> have my doubts: my requirements seem to be at the absolute limits of
> what they offer -- 8/8, lots of fine-pitch high pin count SMDs, and so
> forth -- so I'm not sure whether the results will be any good. Still, it
> doesn't cost much to find out!

I use Mega Electronics JetStar film. My UV exposure is 13 minutes (from 
cold) at about 13 cm above two UV tubes. You might be too close, this can 
cause uneven exposure. I'd rather have a longer exposure with some 
collimation due to distance from the tubes. The PCB material should be quite 
fresh; the resist goes off after about 6 months, this can cause lots of 
problems with resist not being removed during development.

Try 10/10 or 12/12 first, and get that working before you try 8/8. You could 
send me one of your transparencies to try, or I could send you one of mine.

I prefer PCB-Pool to Olimex, their quality is much better and they can do 6 
mil tracks.

Leon

>I've made a number of test boards containing designs with 8/8 rules -
>-in each case the results haven't been nearly good enough to be 
>usable. 

A magnification tool (visor, lamp, glass, loupe, microscope etc..) 
should go a long way in troubleshooting the problem. Do the following;

1. Examine the printed artwork "nine ways to Sunday" using the above 
tool. If there are issues here then it's pointless proceeding with 
the exposure step.

2. Examine the developed board as above. If there are issues here 
then it's pointless proceeding with the etching step.

3. Examine the etched board as above.

Once you've determined where the problem lies then you can focus on 
its correction.


--- In Homebrew_PCBs@yahoogroups.com, Stuart Wallace <stuartw@...> 
wrote:
>

> Leon wrote:
> <Stuff about jetstar>

>> And Stuart wrote:
>> <Stuff about trannys and inkjets>

Guys,  I followed this thread as far up as
I can and I didn't see any mention of ink
being used for the photo tool.

I am only going off hear-say as I use a
tidy laser printer for my photo tools

BUT

Stuart - have you read in the archives that
yellow ink is better for the photo tool
(more UV absorbant) than black ink.  Some
people claimed that black ink was almost
invisable to the UV light.  Though I guess
how "invisable" black is depends on the ink
make up.

Leon - Do you use black ink or a colour ink
for your phototools on the JetStar ?

>> Stuart also wrote:
>> <snip>
>> Something is definitely badly wrong with
>> my process! I struggle to fully develop
>> the boards without mechanical assistance
>> (generally rubbing at the boards with
>> cotton buds) -- this invariably leaves
>> contaminants on the board, which in turn
>> causes etching problems. Etching is in any
>> case very uneven. This seems to point to
>> either bad artwork or insufficient
>> exposure.

Does the bad etching areas look like shiny
clean copper that just wont eat away - or is
it going pinkish but just taking too long ?

If it is still looking "shiny" then you
deffinatly have underexposed/underdeveloped
the mask.

Have you tried doing a "test strip" board ?

>> <snip>
>> I'd love to be able to make 8/8
>> double-sided boards at home; the reduced
>> turnaround time would really help some
>> projects.

You will Dorothy you will.


>> I suspect, however, that DIY PTH may be
>> beyond my abilities.
>> <snip>

PTH is probably beyond my abilities too.

I am not going to let that stop me trying
though.  I just need some more spare time :(

----- Original Message ----- 

From: "Andrew" <andrewm1973@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Monday, June 19, 2006 10:58 PM
Subject: [Homebrew_PCBs] UV Exposure Problems - was - Re: My best fine-pitch 
PCB so far


>> Leon wrote:
>> <Stuff about jetstar>
>
>>> And Stuart wrote:
>>> <Stuff about trannys and inkjets>
>
> Guys,  I followed this thread as far up as
> I can and I didn't see any mention of ink
> being used for the photo tool.
>
> I am only going off hear-say as I use a
> tidy laser printer for my photo tools
>
> BUT
>
> Stuart - have you read in the archives that
> yellow ink is better for the photo tool
> (more UV absorbant) than black ink.  Some
> people claimed that black ink was almost
> invisable to the UV light.  Though I guess
> how "invisable" black is depends on the ink
> make up.
>
> Leon - Do you use black ink or a colour ink
> for your phototools on the JetStar ?

I use black. I'll try yellow and see if it is any better.

Leon

> Leon wrote; 
> I use black. I'll try yellow and see
> if it is any better.

did some more reading - looks like the
difference is more to do with dye and
pigment inks than the colour.

As I said though - I don't use an inkjet
so can't make any real useful comments
and am just repeating what others have
said.

I found that to get anything close to 8mil (0.2mm) features, you need a 
much better collimated light than an array of fluorescent 'BL' tubes. 
Try a 250W clear mercury vapor HID. With correct setup your limitation 
should the capabilities of the inkjet printer/transparency film. My 
Epson Stylus with Epson dye black ink and Epson transparency, you can 
achieve 8mils quite reliably. My Cannon i550 with HP, Kodak or Epson 
transparency film, and using pigment black ink, the reliable minimum 
trace/space width is more like 10mil (0.25mm). Also, the Epson prints 
are so dark, that a 3:1 exposure ratio variation does almost no harm.
I have not tried IBM transparencies. The HP transparency seem to work 
best with pigment ink. Inkjet transparencies in the office are becoming 
increasingly rare because these days people prefer to use digital 
projectors for presentations.

With currently technology, the limitation is in the inkjet printer, and 
how precisely it can place  drops of ink. The DPI claims are absolutely 
meaningless for PCB artwork if the drops can wonder off a few thousands 
of an inch. Look at your transparency prints under a microscope and you 
will see what I mean. There are no market expectations to encourage 
inkjet manufactures to build printers with more precisely placed drops. 
This is a shame because the potential is there.


Adam

Stuart Wallace wrote:

> Leon Heller wrote:
> 
> 
>>What are you printing the transparencies on?
>> 
>>
> 
> I'm using a Canon i560 to print on IBM inkjet transparencies. I've tried 
> a few different types of easily-available transparencies (the sort of 
> stuff that places like PC World stock) and haven't noticed much of a 
> difference between the types. I've made a number of test boards 
> containing designs with 8/8 rules -- in each case the results haven't 
> been nearly good enough to be usable. Something is definitely badly 
> wrong with my process! I struggle to fully develop the boards without 
> mechanical assistance (generally rubbing at the boards with cotton buds) 
> -- this invariably leaves contaminants on the board, which in turn 
> causes etching problems. Etching is in any case very uneven. This seems 
> to point to either bad artwork or insufficient exposure.
> 
> I expose the boards (6" x 4" DS FR4 stock) about two inches above two 
> 15W UV tubes for three minutes. I'm extremely nervous about 
> over-exposure: perhaps too nervous!
> 
> I'd love to be able to make 8/8 double-sided boards at home; the reduced 
> turnaround time would really help some projects. I suspect, however, 
> that DIY PTH may be beyond my abilities. I'm getting ready to send 
> artwork over to Olimex in order to get some proto boards made up, but I 
> have my doubts: my requirements seem to be at the absolute limits of 
> what they offer -- 8/8, lots of fine-pitch high pin count SMDs, and so 
> forth -- so I'm not sure whether the results will be any good. Still, it 
> doesn't cost much to find out!
> 
> 
> Stuart
> 
> 
> 
> Be sure to visit the group home and check for new Links, Files, and Photos:
> http://groups.yahoo.com/group/Homebrew_PCBs
> 
> If Files or Photos are running short of space, post them here:
> http://groups.yahoo.com/group/Homebrew_PCBs_Archives/ 
> Yahoo! Groups Links
> 
> 
> 
>  
> 
> 
> 
>

>I found that to get anything close to 8mil (0.2mm) features, you 
>need a much better collimated light than an array of 
>fluorescent 'BL' tubes.

8 mil is a walk in the park with or without collimation provided 
you're using a contact frame. Less than 5 mil may benefit from it but 
by that time you've hit the resolution limitations of the inkjet 
printer.

--- In Homebrew_PCBs@yahoogroups.com, Adam Seychell <a_seychell@...> 
wrote:
>

>> Adam Seychell wrote:
>> I found that to get anything close to 8mil
>> (0.2mm) features, you need a much better
>> collimated light than an array of fluorescent
>> 'BL' tubes.

> derekhawkins wrote: 
> 8 mil is a walk in the park with or without
> collimation provided you're using a contact
> frame. Less than 5 mil may benefit from it but 
> by that time you've hit the resolution
> limitations of the inkjet printer.


I agree that 8/8 is quite do-able with a contact
frame and UV tubes.

However - what I am calling 8/8 and what others
call 8/8 may differ.

I think that I have sucessfully made an 8/8
board if I can get 150mm x 300mm pre-coated PCB
to have no shorts and no breaks across a dense
area of 8 thou tracks and 8 thou spaces AND
still be able to see that with my luxo desk lamp
that the lines all look like they have straight
edges and that the track thickness does not vary
to greatly. (i dont use a microscope just a
magnifier)

Some people might claim that the optical and
chemical undercut that makes my 8 thou tracks
realy be 6 or 7 thou in places be a failure.

It's all up to what you call a success i guess.

I do agree collimated light is a big advantage.

I am going to build a good collimated source
one day.  But for now I think I get reliable
8/8 from a Kinsten suitcase exposure box.

I think better results will come from the
photo plotter (which is still slowly moving
along).  Hopefully I will move to the 6/6
realm when it is done (maybe even 4/6).

Andrew wrote:
> 
> I think better results will come from the
> photo plotter (which is still slowly moving
> along).  Hopefully I will move to the 6/6
> realm when it is done (maybe even 4/6).

Are you actually building your own photoplotter ? I'm interested how 
your doing it.
Is it a rotating drum with a slow scan single pass laser ?
Or is it the classic X-Y table with a light pen ?
What are you using for the photosensitive film, and what light color 
sensitivity does it have.

The red light sensitive films normally used in photoplotters can only be 
handled under very dim green light.

> Andrew wrote:
> > 
> > I think better results will come from the
> > photo plotter (which is still slowly moving
> > along).  Hopefully I will move to the 6/6
> > realm when it is done (maybe even 4/6).

> Adam S wrote:
> 
> Are you actually building your own
> photoplotter ? I'm interested how your doing
> it.  Is it a rotating drum with a slow scan
> single pass laser ?

Yes - Rotating drum. Single Laser. 15 minutes
a page kinda thing.  Just a rehash of the one
John Elson (pico systems) did.

I hope to make it cheap and easy enough for
anyone to make with mainly parts you could get
at places like Jaycar. (cheap and easy are
relative terms I know :D)

Original plan to make it computer system
independant hit some snags and have since
spent a few weeks learning how to write
windows applications (win32 API).  It is
coming along very slowly as it also has to
share my time with other projects.

My original estimate was 6 months to finish
and I think I can still make that :D

> Or is it the classic X-Y table with a light
> pen ?

God no - that sounds WAY too hard.

> What are you using for the photosensitive
> film, and what light color sensitivity
> does it have.

Red sensitive photoplotter film.  It has
"dual sensitivity" at 635 and 680 nm I think.
Whatever the wavelengths are for common HeNe
and solid state lasers are.

I think it works out something like $5-6 for
a sheet a bit bigger than A3 size.  I can't
quite remember exactly though - and I don't
have the details here.

> The red light sensitive films normally used
> in photoplotters can only be handled under
> very dim green light.

Yes - green light safe lights are the go.

I have not found out how sensitve it is to
stray light yet - as I am waiting till I have
finished the scanner before I buy any of the
film.

Does anyone have some spare time they can
lend me :D

Andrew wrote:
> Yes - Rotating drum. Single Laser. 15 minutes
> a page kinda thing.  Just a rehash of the one
> John Elson (pico systems) did.

Sounds like fun. I'd like to have a go at building something like that myself. 
Could be kinda neat to play with photo-etching.

Apparently you can use UV LEDs to expose PCB photoresist - there was an 
article in Elektor about it. They modified a plastic storage bin by adding a 
PCB with some UV LEDs to the bottom, then putting a plastic (or maybe it was 
glass) sheet on top.
I'd be tempted to get a bit of MDF and make a custom box for it, then rig it 
to expose both sides at once.

> I hope to make it cheap and easy enough for
> anyone to make with mainly parts you could get
> at places like Jaycar. (cheap and easy are
> relative terms I know :D)

The hard part is going to be finding the steppers, and focussing the laser 
down enough to get a decent amount of resolution out of the film. I haven't 
seen any laser pointers that can be focussed down below 1mm...

> It is
> coming along very slowly as it also has to
> share my time with other projects.

Heh, I know how that goes.. I've got a battery charger on my workbench waiting 
for some serious debugging work to be done on it. I need to smooth out the 
current sense line to try and get the regulator to stabilise. Switchmode PSUs 
are neat, but annoyingly difficult to get working reliably.

Even bought a pack of Energiser 2500mAh AAs to play with (but I'm doing the 
development work with a Kodak 1600mAh that seems to have lost its mate). Both 
the Energisers and the Kodaks are rebranded Sanyo cells, to the best of my 
knowledge. Easy way to tell (on the namebrand cells anyway) is to look for the 
text "-H-R-" stamped into the negative contact on the battery. That only 
applies to name-brand cells though (Duracell and Energiser mainly) - a lot of 
unscrupulous Chinese manufacturers are doing their best to clone the Sanyo 
cells, as cheaply as possible. Which generally means the "2500mAh" clones are 
only good to about 1500mAh at best, and the delta-peak isn't as pronounced, so 
the chargers tend to miss it and end up overcharging the cells.

>> Or is it the classic X-Y table with a light
>> pen ?
> 
> God no - that sounds WAY too hard.

XY isn't that hard - I've got a few scribbly designs and CAD drawings here. 
What is hard is taking XY and adding Z axis control too (i.e. for a CNC drill).
I want to make a CNC drill, but I need to get my mitts on another pair of 
stepper motors (preferably Laserjet II or III steppers, so I've got a matched 
pair) and a decent power supply. Then I need to track down some nice, smooth 
slides and a leadscrew (or something similar) to move the tracks around...

> Red sensitive photoplotter film.  It has
> "dual sensitivity" at 635 and 680 nm I think.
> Whatever the wavelengths are for common HeNe
> and solid state lasers are.

Where do you get that stuff from? I've never seen it for sale anywhere...
How do you develop it anyway (in fact, do you need to develop it at all)? I 
did a bit of B&W photography a while ago, so I'm thinking of this from a "how 
is photoplotter film different to B&W film/paper emulsion" perspective...

> I think it works out something like $5-6 for
> a sheet a bit bigger than A3 size.  I can't
> quite remember exactly though - and I don't
> have the details here.

So that's what, $2.50 an A4 sheet, or $1 for the average Eurocard sized PCB?

> Does anyone have some spare time they can
> lend me :D

Sorry, I let someone borrow my last Round Tuit and they still haven't returned 
it. In addition, my supplier of Round Tuits has been out of stock for the past 
ten years :)

Thanks.
-- 
Phil.                         | Kitsune: Acorn RiscPC SA202 64M+6G ViewFinder
philpem@...         | Cheetah: Athlon64 3200+ A8VDeluxeV2 512M+100G
http://www.philpem.me.uk/     | Tiger: Toshiba SatPro4600 Celeron700 256M+40G

--- In Homebrew_PCBs@yahoogroups.com, Adam Seychell <a_seychell@...> 
wrote:
>
> 
> With currently technology, the limitation is in the inkjet printer, 
and 
> how precisely it can place  drops of ink. The DPI claims are 
absolutely 
> meaningless for PCB artwork if the drops can wonder off a few 
thousands 
> of an inch. Look at your transparency prints under a microscope and 
you 
> will see what I mean. There are no market expectations to encourage 
> inkjet manufactures to build printers with more precisely placed 
drops. 
> This is a shame because the potential is there.
> 
> 
> Adam
> 
> > > 
> > 

First off I apologize if I am taking the above too far out of 
context. It is not my intention to take the statements out of 
context, only to point out that the DPI limitations are NOT STRICTLY 
RELATED TO THE PRINTER ALONE. I have done many experiments with 
transparency materials for use with photo-etching and can say that in 
MY experiences the various TRANPARENCY MATERIALS fluctuate greatly 
between brands and batches of the same brand. I  find this 
fluctuation in both the ability of the material to quickly absorb ink 
and the Coarseness ( unevenness or GRAIN ) of the side intended to 
receive the ink. Stop and consider what the intended purpose of MOST 
of these materials ultimately is, OVERHEAD PROJECTION at very high 
magnification, not a lot of requirement for high resolution.

Using an OLD Epson STYLUS 1160 (supposed resolution of up to 720 x 
1440 and standard Epson ink) The best material I have found is that 
which is INTENDED for the creation of  "contact negatives" for 
silkscreen printing.

I have posted this here before I believe, but as the subject comes up 
now I'll post again, 
< http://screenprinters.net/product_group.php?gid=inkjetfilm >
At one time you could request a free SAMPLE pack of the material. I 
don't know if that offer still stands but you could search the site 
for the offer or send an email requesting a sample and you might get 
it.

>I agree that 8/8 is quite do-able with a contact
>frame and UV tubes.

6/6 is doable here but I've never had reason to with anything real. 
Here is tortuous 6/6 artwork, this photo doesn't do it justice due to 
moire and other artifacts;

http://www.pbase.com/eldata/image/62158168

Here's a closeup of a section;

http://www.pbase.com/eldata/image/62158260

Basically, it's 150 concentric rectangles resulting in 300 parallel 
horizontal and vertical 6 mil tracks with 6 mil spacing. I can do a 
faithful reproduction in copper without collimation which leads me to 
believe the need for collimation at these pedestrian resolutions is a 
myth.




--- In Homebrew_PCBs@yahoogroups.com, "Andrew" <andrewm1973@...> 
wrote:
>

>> andrewm wrote:
>> I agree that 8/8 is quite do-able with a
>> contact frame and UV tubes.

> DerekH wrote:
> 6/6 is doable here but I've never had
> reason to with anything real. Here is
> tortuous 6/6 artwork, this photo doesn't
> do it justice due to moire and other
> artifacts;
> 
> http://www.pbase.com/eldata/image/62158168
> 
> Here's a closeup of a section;
> 
> http://www.pbase.com/eldata/image/62158260
> 
> Basically, it's 150 concentric rectangles
> resulting in 300 parallel horizontal and
> vertical 6 mil tracks with 6 mil spacing.
> I can do a faithful reproduction in copper
> without collimation which leads me to 
> believe the need for collimation at these
> pedestrian resolutions is a myth.

Derek - what are you using as a phototool
to get good 6/6 ?

I can get 6/6 when I borrow a fancy laser
printer to print on vellum.  However my
personal laser printer at home (HP Laserjet
4xxx) does not give me enough contrast to
get realy reliable mask down there.

In fact the difference between the car
priced 1200 dpi colour laser I borrow and
my 1200dpi monochrome HP is quite amazing.
With my HP I can do 8/8 tracks and have
them look pretty good but large copper
areas get pinholes and washouts in them.

With the borrowed printer large areas of
black stay perfectly black and you can see
2 pixel wide (600dpi) jaggies on diaganol
lines after etching.  8/8 looks perfect
and you can't see any rough edges to the
lines at all with a desk magnifier.

> DerekH wrote:

> <snip losts>
> Here's a closeup of a section;
> 
> http://www.pbase.com/eldata/image/62158260
> <snip lots more>

Derek - Like your closeups of the SMD
components.

I am looking at getting a macro lens for my
Nikon D70 as I miss not having a macro from
my old CP5700.

Your photos of the SMD parts say 100mm.

Is that lens a 100mm in 35mm terms or is
that 100mm after the digital SLR one point
whatever multiplication factor ?

I was looking at a 105mm macro lens that
would end up being equivilent to something
like 140mm on the digital and was wondering
if that was too much.  I was used to a 90mm
on my old pentax film cameras and would not
have minded a bit longer back then.

Philip Pemberton wrote:
> 
>>I hope to make it cheap and easy enough for
>>anyone to make with mainly parts you could get
>>at places like Jaycar. (cheap and easy are
>>relative terms I know :D)
> 
> 
> The hard part is going to be finding the steppers, and focussing the laser 
> down enough to get a decent amount of resolution out of the film. I haven't 
> seen any laser pointers that can be focussed down below 1mm...
> 

Yes, cheap and easy are the two words I had in mind when pondering on 
the idea of a homebrewed photoplotter. The dedicated photoplotter film 
was a bit pricey from memory, but worse was the $1000 odd minimum order 
quantity for me (in Australia). I didn't shop around  and I cant 
remember who I called, may be it was Kodak ?

http://www.kodak.com/eknec/PageQuerier.jhtml?pq-path=2608/2610/4131&pq-locale=en_US




> Where do you get that stuff from? I've never seen it for sale anywhere...
> How do you develop it anyway (in fact, do you need to develop it at all)? I 
> did a bit of B&W photography a while ago, so I'm thinking of this from a "how 
> is photoplotter film different to B&W film/paper emulsion" perspective...

I believe the film is processed just as B&W film. I don't know what the 
difference is with B&W film and 'photoplotter' film besides the spectral 
sensitivity. Maybe B&W can work in a photoplotter, with appropriate 
laser.  It seems lasers other than red are very expensive so that may be 
their limitation.

The most accurate drive mechanism will be a stepper motor and a toothed 
belt drive combined with a linear optical encoder. Better than 20um 
resolution should be possible.

>Derek - what are you using as a phototool
>to get good 6/6 ?

Epson R320 and Epson transparencies. I have two HP lasers here and 
used them until the dimensional inaccuracies started to get in the way 
of CNC drilling. Actually used laser transparencies instead of vellum 
so was up to my neck with pinhole correction and solvent fumigation 
tricks. My exposure box consists of normal daylight fluorescent 
instead of UV tubes since it had to double as a standard light box for 
pinhole correction. 

--- In Homebrew_PCBs@yahoogroups.com, "Andrew" <andrewm1973@...> wrote:
>

> AdamS wrote:
>
> Yes, cheap and easy are the two words I
> had in mind when pondering on the idea
> of a homebrewed photoplotter. The
> dedicated photoplotter film was a bit
> pricey from memory, but worse was the
> $1000 odd minimum order quantity for me
> (in Australia). I didn't shop around 
> and I cant remember who I called, may
> be it was Kodak ?

>http://www.kodak.com/eknec/PageQuerier.
>jhtml?pq-path=2608/2610/4131&pq-locale=
>en_US

Yep RED7 was the film I was looking at.

And the 1000$ AUD is a bit of a steep
price of entry.  But I Think it might
be worth it even if you can't find a
friendly person to give it to you 10
sheets at a time.

It's just comical that the cost of
consumables will exceed the cost of
hardware 10:1


> The most accurate drive mechanism
> will be a stepper motor and a toothed 
> belt drive combined with a linear
> optical encoder. Better than 20um 
> resolution should be possible.

Yep - have all those parts sitting in
the garage just waiting.

My only concern is if sticktion is
going to cause a problem.

Time will tell.

> PhilipP wrote:

> <snip>
> Apparently you can use UV LEDs to expose
> PCB photoresist - there was an  article
> in Elektor about it. They modified a
> plastic storage bin by adding a PCB with
> some UV LEDs to the bottom
> <snip>

UV tubes are still pretty cheap compared
to UV leads if you are going to make an
exposure box that is more than about 100
mm x 100mm.

> > I hope to make it cheap and easy enough for
> > anyone to make with mainly parts you could get
> > at places like Jaycar. (cheap and easy are
> > relative terms I know :D)
> 
> The hard part is going to be finding the
> steppers,

DC servo motor for drum.
Off the shelf stepper and a belt for the y axis
is the present plan

> and focussing the laser down enough to get a
> decent amount of resolution out of the film. I
> haven't seen any laser pointers that can be
> focussed down below 1mm...

I have found a supplier of lasers with focusing
optics that claim 10 or 12 uM and I will use
these for early prototypes.  However I would
like to do an easy to make modification to
common laser pointers with a bit of alfoil
for others to be able to make.

However even if I can't - the comercial 12uM
ones where sub $200 AUD.  Which is twice my
ideal BOM for the ENTIRE final unit - but
still bearable.

> <snip>
>> Red sensitive photoplotter film. It has
>> "dual sensitivity" at 635 and 680 nm I think.
>> Whatever the wavelengths are for common HeNe
>> and solid state lasers are.

> Where do you get that stuff from? I've never
> seen it for sale anywhere...

It's not real easy to get.

First you have to get some chalk and draw a
pentagram on the floor.  Next you have to
prepare the goat......

> How do you develop it anyway (in fact, do
> you need to develop it at all)?

Two step process like BW film for developing.

Develop then Fix.

The developer and fixer is specially designed
for the film.  The deamon that hands you the
blood filled pen when you get the film should
also be able to supply the developer/fixer.

> I did a bit of B&W photography a while ago,
> so I'm thinking of this from a "how is
> photoplotter film different to B&W film/paper
> emulsion" perspective...

Well - it's clear and black rather than white
and black.  It also has obscene contrast levels.
If you look at the sensitivity plots it is very
very non-linear.  Goes very rapidly from clear
to black with no grey shades in between.

The other difference is the lack of market
for it - so the suppliers can be as difficult
as they like and charge what ever price they
want.

Though with epson now making 4 shades of grey
ink jet printers I think that the market for
black and white paper is about to get about as
thin.

> > I think it works out something like $5-6 for
> > a sheet a bit bigger than A3 size.  I can't
> > quite remember exactly though - and I don't
> > have the details here.
> 
> So that's what, $2.50 an A4 sheet, or $1 for
> the average Eurocard sized PCB?
> 

I think so - I will have to check the prices
though.  Will get back to you all on the week-
end when I have checked.

> PhilipP wrote:

> <snip>
> Apparently you can use UV LEDs to expose
> PCB photoresist - there was an  article
> in Elektor about it. They modified a
> plastic storage bin by adding a PCB with
> some UV LEDs to the bottom
> <snip>

UV tubes are still pretty cheap compared
to UV leads if you are going to make an
exposure box that is more than about 100
mm x 100mm.

> > I hope to make it cheap and easy enough for
> > anyone to make with mainly parts you could get
> > at places like Jaycar. (cheap and easy are
> > relative terms I know :D)
> 
> The hard part is going to be finding the
> steppers,

DC servo motor for drum.
Off the shelf stepper and a belt for the y axis
is the present plan

> and focussing the laser down enough to get a
> decent amount of resolution out of the film. I
> haven't seen any laser pointers that can be
> focussed down below 1mm...

I have found a supplier of lasers with focusing
optics that claim 10 or 12 uM and I will use
these for early prototypes.  However I would
like to do an easy to make modification to
common laser pointers with a bit of alfoil
for others to be able to make.

However even if I can't - the comercial 12uM
ones where sub $200 AUD.  Which is twice my
ideal BOM for the ENTIRE final unit - but
still bearable.

> <snip>
>> Red sensitive photoplotter film. It has
>> "dual sensitivity" at 635 and 680 nm I think.
>> Whatever the wavelengths are for common HeNe
>> and solid state lasers are.

> Where do you get that stuff from? I've never
> seen it for sale anywhere...

It's not real easy to get.

First you have to get some chalk and draw a
pentagram on the floor.  Next you have to
prepare the goat......

> How do you develop it anyway (in fact, do
> you need to develop it at all)?

Two step process like BW film for developing.

Develop then Fix.

The developer and fixer is specially designed
for the film.  The deamon that hands you the
blood filled pen when you get the film should
also be able to supply the developer/fixer.

> I did a bit of B&W photography a while ago,
> so I'm thinking of this from a "how is
> photoplotter film different to B&W film/paper
> emulsion" perspective...

Well - it's clear and black rather than white
and black.  It also has obscene contrast levels.
If you look at the sensitivity plots it is very
very non-linear.  Goes very rapidly from clear
to black with no grey shades in between.

The other difference is the lack of market
for it - so the suppliers can be as difficult
as they like and charge what ever price they
want.

Though with epson now making 4 shades of grey
ink jet printers I think that the market for
black and white paper is about to get about as
thin.

> > I think it works out something like $5-6 for
> > a sheet a bit bigger than A3 size.  I can't
> > quite remember exactly though - and I don't
> > have the details here.
> 
> So that's what, $2.50 an A4 sheet, or $1 for
> the average Eurocard sized PCB?
> 

I think so - I will have to check the prices
though.  Will get back to you all on the week-
end when I have checked.

>I am looking at getting a macro lens for my
>Nikon D70 as I miss not having a macro from
>my old CP5700.

I was in the Nikon camp too once, have some CP5700 shots here;

http://www.pbase.com/eldata/cp5700

>Is that lens a 100mm in 35mm terms or is
>that 100mm after the digital SLR one point
>whatever multiplication factor ?

100mm in 35mm terms.

>I was looking at a 105mm macro lens that
>would end up being equivilent to something
>like 140mm on the digital and was wondering
>if that was too much. 

Not at all. For a close macro shot you'll be about a foot from the 
subject instead of a few inches in the case of say a CP5700. But 
close macro shots have a downside in terms of DOF as you probably 
know. Often it's better not to get too close then do a 100% crop in 
your photo editor.

--- In Homebrew_PCBs@yahoogroups.com, "Andrew" <andrewm1973@...> 
wrote:
>

>> Me wrote:
>> I was looking at a 105mm macro lens that
>> would end up being equivilent to something
>> like 140mm on the digital and was wondering
>> if that was too much. 

> Derek wrote:
>
> Not at all. For a close macro shot you'll be
> about a foot from the subject instead of a
> few inches in the case of say a CP5700. But 
> close macro shots have a downside in terms
> of DOF as you probably know. Often it's
> better not to get too close then do a 100%
> crop in  your photo editor.

Though your butterflies are pretty I don't have
that much patience.  All my macro photography
will be pretty stationry (PCBs, gears, scratches)

I used to get around the DOF problem with
my old pentax film camera by just shutting
down to F22 and leaving it on a tripod for a
while.

I found the digital cameras (like my old
CP5700) with their tiny sensors used to get
good DOF regardless of apature.  I am hoping
the small sensor in the D70 might help a bit
as well.

>I used to get around the DOF problem with
>my old pentax film camera by just shutting
>down to F22 and leaving it on a tripod for a
>while.

I take that approach too when necessary but there are at least two 
downsides with digital that this F22 shot demonstrates (used flash 
instead of a tripod);

http://www.pbase.com/eldata/image/24426072


--- In Homebrew_PCBs@yahoogroups.com, "Andrew" <andrewm1973@...> wrote:
>

>I used to get around the DOF problem with
>my old pentax film camera by just shutting
>down to F22 and leaving it on a tripod for a
>while.

I take that approach too when necessary but there are at least two 
downsides with digital that this F22 shot demonstrates (used flash 
instead of a tripod);

http://www.pbase.com/eldata/image/24426072


--- In Homebrew_PCBs@yahoogroups.com, "Andrew" <andrewm1973@...> wrote:
>